This project aims to distribute take-over-requests (ToR) for automated connected vehicles using reinforcement learning. It used single and multi agent approaches.
This branch contains only the final results (trained models, sim results, plot) with some explanation.
The models order presentation is presented in ModelsOrder.png.
The presentation.pdf contains all the overview of the project.
To test every approach checkout to the sutiable branch and follow the readme file.
For Single Agent approach:
git fetch
git checkout SingleAgent
For Multi Agent approach:
git fetch
git checkout MultiAgent
You can clone also every branch seperately by the following
git clone https://github.com/anoulis/Reinforcement-Learning-in-Traffic-Control-For-Connected-Automated-Vehicles.git -b SingleAgent
git clone https://github.com/anoulis/Reinforcement-Learning-in-Traffic-Control-For-Connected-Automated-Vehicles.git -b MultiAgent